Coupling of the anthracycline antitumor drug menogaril to 2

Qibing Zhou and Kenneth D. Turnbull. The Journal of Organic ... Carleen Cullinane , Anne van Rosmalen , and Don R. Phillips. Biochemistry 1994 33 (15)...
0 downloads 0 Views 406KB Size
J . Am. Chem. SOC.1989, 1 1 1 , 8291-8293 one another in the polypeptide chain, its intactness indicates that no major distortion of the protein’s native, folded structure occurs when the dry enzyme is suspended in anhydrous acetone or octane. Furthermore, there is no evidence of appreciable structural heterogeneity in the population of enzyme active centers in these solvents. Thus, the vast difference in the catalytic efficiency of a-LP in water, acetone, and octane (k,,/K, = 5800, 3.3, and 5400 M-’ m i d , respectively)14is not due to disruption of conformation but rather to another effect of the solvent, e.g., a change in the binding affinities of substrates or in protein dynamics (perhaps by stripping bound In contrast to the above results, dimethyl sulfoxide (DMSO, in which proteins are soluble16)was found to disrupt the catalytic triad. The MAS spectrum of I5N*’-Hisa-LP (“pH 8.6”) reclaimed from DMSO” displayed (Figure lC), in addition to the peptide backbone peak, a single resonance at 120 ppm, indicating the exclusive existence of the NfZ-H tautomereg This is consistent with the observations that (i) DMSO was the only solvent of seven testedI4 in which a-LP was inactive and (ii) a-LP lyophilized from DMSO (as opposed to water) was inactive when suspended in octane. In summary, solid-state MAS N M R spectroscopy has been used to show that the catalytic triad of a - L P is intact in acetone and octane, thus providing the first direct physical evidence that enzymes do not denature when suspended in neat organic solvents. Registry No. His, 71 -00-1 ; a-lytic proteinase, 37288-76-9; acetone, 67-64-1; octane, 11 1-65-9; dimethyl sulfoxide, 67-68-5. (14) a-LP was assayed in organic solvents by following transesterification of N-AC-L-AI~-OCH~CH~CI with ethanol as described for subtili~in’~ and in water by following hydrolysis of the same substrate using a pH-stat. a-LP preinhibited with phenylmethanesulfonyl fluoride was inactive in organic solvents. (15) Zaks, A.; Klibanov, A. M. J . Biol. Chem. 1988, 263, 3194. (16) S k i e r , S . J. Adu. Protein Chem. 1962, 17, 1. (1 7) ISN I-His a-LP (‘pH 8.6”) was dissolved (0.2 mg/mL) in DMSO and concentrated by vacuum distillation and lyophilization. (18) This work was financially supported by NSF Grant CBT-8710106 to A.M.K.

Coupling of the Anthracycline Antitumor Drug Menogaril to 2’-Deoxyguanosine through Reductive Activation’ Michael Egholm and Tad H. Koch* Department of Chemistry and Biochemistry University of Colorado Boulder, Colorado 80309-0215 Received April 24, 1989 The concept of bioreductive activation as applied by Moore and Czerniak to the anthracycline antitumor drugs, most notably daunomycin and adriamycin, proposes formation of a reactive quinone methide from a hydroquinone state.13 The hydroquinone state is formed through bioreduction, and the quinone methide reacts with a nucleophilic site in a biological macromolecule. The tumor cell toxicity is proposed to result from the reaction of the quinone methide with a nucleophilic site in DNA such as the (1) We gratefully acknowledge the financial assistance from PHS in the form of Grant CA-24665 and the NIH Small Instrumentation and Biomedical Sciences Support Grants to the University of Colorado (DHHS) and gifts of menogaril and 7-deoxynogarol from the Upjohn Co., courtesy of Paul Wiley. M.E. thanks the Danish Research Academy for partial support, and T.H.K. thanks the University of Colorado Council on Research and Creative Work for a faculty fellowship. We thank Martin Ashley for assistance with the NMR measurements, Ronald Sadecky and Robert Barkley for assistance with the mass spectral measurements, and Giorgio Gaudiano and Makhluf Haddadin for helpful discussions. (2) Moore, H. W. Science (Washington, D.C.) 1977, 197, 527. Moore, H. W.; Czerniak, R. Med. Res. Rev. 1981, I, 249. (3) Abdella, B. R. J.; Fisher, J. EHP, Enuiron. Health Perspecr. 1985.64, 3. Powis, G. Pharmacol. Ther. 1987, 35, 57.

8291

2-amino group of 2’-deoxyguanosine. Although such a coupling in itself may not be cytotoxic, it attaches an effective catalyst for the production of reactive oxygen species to DNA. Subsequent catalytic production and reaction of reactive oxygen species may damage the DNA beyond repair.4 Reductive activation of the quinone antitumor drug mitomycin C leads to coupling to guanine m ~ i e t i e s and , ~ the elegant work of Tomasz and Nakanishi has further established that reductively activated mitomycin C cross-links duplex DNA through reaction with two guanine bases.6 However, Fisher and Aristoff have recently stated in their extensive review of the chemistry of DNA modification by antitumor antibiotics with regard to the quinone methide from reductive activation of anthracyclines: ”in principle this quinone methide may be thought of as a DNA-reactive entity, but its behavior with DNA on this score in vitro is ambiguous and numerous efforts to obtain mononucleotide adducts have been unsuccessful” and further “there is no in vitro experiment to provide even the suggestion of the chemistry leading to adduct formation”.’ We have recently reported that reduction of the anthracycline antitumor drug menogaril (1) with substoichiometrk quantities of the one-electron-reducing agent bi(3,5-dimethyl-5-(hydroxymethyl)-2-oxomorpholin-3-yl) (2, DHM-3 dimer) in an aqueous medium yields a long-lived quinone methide transient, which can be trapped with the sulfur and nitrogen nucleophiles N-acetylcysteine* and imidazole? respectively. These results suggest that the quinone methide can react with cysteine and histidine residues in peptides and proteins. We now describe the successful reductive coupling of menogaril at the 7-position to 2’-deoxyguanosine at the 2-amino position. Menogaril (1 8 mg, 2.0 X 1O4 M) was reacted anaerobically with 25 mol % of DHM-3 dimer (2) in 10% dimethylformamide-90% pH 8.2, aqueous, Tris buffer (v/v) in the presence of a 100-fold excess of 2’-deoxyguanosine for 6 days at ambient temperature. Reverse-phase HPLC analysis of the reaction mixture eluting with a gradient consisting of a formate buffer and methanol and detecting at 480 nm showed peaks with retention times of 1.7 min with a shoulder at 2.0 min (1 7%, 7-(2”-N(2’”-deoxyguanosyl))-7-deoxynogarol stereoisomers,I0 3), 5.3 min (17%, menogaril, l), 5.6 min (30%, 7-deoxynogarol, 4), 6.8 min (23%, bi(7-deoxynogaro1-7-y1), 5 ) , and 9.7 min (lo%, probably a stereoisomer of 5). Minor unidentified product peaks appeared at 2.5 min (2%) and 2.8 min (1%). Menogaril, 7-deoxynogarol, and the major stereoisomer of bi( 7-deoxynogarol-7-yl) were identified by coinjection onto the HPLC column with authentic samplesa and purity evaluation of the peaks with a HewlettPackard diode array detector. The deoxynogarol dimer 5 was further characterized by isolation from a chloroform extract of the reaction mixture and FAB mass spectral analysis, which showed a characteristic molecular ion at m / z 1021.* The major stereoisomer of the deoxyguanosyl-deoxynogarol adduct 3 was isolated from the aqueous phase after the chloroform extraction by evaporation of the water followed by a series of ethanol extractions of the solid material to remove excess deoxyguanosine and chloroform extractions of aqueous solutions to remove byproducts. Extraction was followed by reverse-phase flash chro(4) Bachur, N. R.; Gordon, S. L.; Gee, M. V. Mol. Pharmacol. 1977, 13, 901. Lown, J. W.; Sim, S.-K.; Majumdar, K. C.; Change, R. Y. Biochem. Biophys. Res. Commun. 1977,76,706. Bachur, N. R.; Gee,M. V.; Friedman, R. D. Cancer Res. 1982,42, 1078. Doroshow, J. H. Proc. Narl. Acad. Sci. U.S.A. 1986, 83, 4514. (5) Hashimoto, Y.; Shudo, K.; Okamoto, T. Tetrahedron Lett. 1982, 23, 677. Tomasz, M.; Lipman, R.; Snyder, J. K.; Nakanishi, K. J . Am. Chem. SOC.1983, 105, 2059. (6) Tomasz, M.; Lipman, R.; McGuinness, B. F.; Nakanishi, K. J . Am. Chem. SOC. 1988, 110, 5892. Tomasz, M.; Lipman, R.; Chowdary, D.; Pawlak, J.; Verdine, G. L.; Nakanishi, K. Science (Washington, D.C.) 1987, 235, 1204. (7) Fisher, J. F.; Aristoff, P. A. Prog. Drug Res. 1988, 32, 411. (8) Boldt, M.; Gaudiano, G.; Haddadin, M. J.; Koch, T. H. J. Am. Chem. SOC.1989, 1 1 1 , 2283. (9) Gaudiano, G.; Egholm, M.; Haddadin, M. J.; Koch, T. H. J. Org. Chem., in press. (10) The minor product was not isolated and is presumed to be a diastereomer of 3 on the basis of its similar chromatographic behavior.

0002-786318911511-8291$01.50/0 0 1989 American Chemical Society

8292 J . Am. Chem. SOC., Vol. 111, No. 21, 1989

Communications to the Editor

Scheme I

DHM-3 Dimer 2 -

DHM-3

7-Deoxynoparol Cntniyels

4_

Menopsrll Menopnrii

+

I ,1

Sml-

qulnona

DHM-3

I

.‘

I

I

Menopsril

,“:p,b“,‘e +

B Quinone Melhlde

-6 Menoparii

+

Menoparii Hydroquinone

HO 2’-deoxyguanorlna

:

HO

2”’

HO

matography eluting with 40% of 0.1% trifluoroacetic acid-60% methanol (v/v). A major fraction from the column contained one stereoisomer of 3 contaminated with about 10% of the other isomer and only 4% guanosine (peak area percent) as indicated by HPLC analysis detecting at 260 nm. The structure of the major isomer isolated was established by a characteristic menogaril quinone UV-vis spectrum, by a strong FAB M 1 mass spectral peak at m/z 777 with a fragment ion at m/z 645 corresponding to loss of the deoxyribose group, and by one- and two-dimensional IH N M R spectroscopy. The one-dimensional 500-MHz IH N M R spectrum of 3 in DMSO-d, solvent was assigned as follows: 6 12.35 and 12.20 (s, 4- and 6-OH), 10.66 (s, l”-NH), 9.01 (br s, 3’-N+HMez), 7.95 (s, S”), 7.42 (s, l l ) , 7.16 (s, 3), 6.67 (d, J = 8 Hz, 2”-NH), 6.21 (t, J = 7 Hz, l”’), 5.74 (d, J = 4 Hz, 1’), 5.56 (m, 7), 5.2 (br, aliphatic OH’S), 4.4 (m, 3”’), 4.19 (dd, J = 4, 10 Hz, 2’), 3.99 (d, J = 10 Hz, 4’), 3.6-3.85 (m, 5’9, 3.04 (d, J = 18 Hz, 10 (a orb)), 2.88 (s, 3’-NMe,), 2.83 and 2.85 (s, 3’-N+HMe2),I12.7-2.9 (m, 10 (a or b), 3’, 2’” (a or b)), 2.23 (m, 2”’ (a orb)), 1.9-2.1 (m, 8 (a and b)), 1.64, (s, 5’-CH3), and 1.32 (s, 9-CHJ. A double quantum filtered 500-MHz homonuclear COSY spectrum established connectivity between the 7-position and the 2”-NH position. The assignment of the 2”-NH was further established by D20 exchange. The COSY spectrum and the spectrum with deuterium exchange also facilitated the chemical shift assignments as reported above, particularly when patterns overlapped such as those for protons at positions 10, 3’, and 2”’. Formation of 7-(2”-N-(2”’-deoxyguanosyl))-7-deoxynogarol (3), 7-deoxynogarol (4), and bi(7-deoxynogarol-7-yl) (5) is proposed to occur via the quinone methide intermediate 6 as shown in Scheme I. Tautomerization of 6, catalyzed by any hydroquinone present such as from reduction of starting material or product, gives 7-deoxynogaroL8 Dimerization of 6 gives initially 5 in the half quinone-half hydroquinone state. Nucleophilic addition of the 2’-NH2 of guanosine to the 7-position of 6 followed by proton shifts gives adduct 3 in its hydroquinone state 7. The

nucleophilic addition and dimerization steps are possibly rever~ i b l e . ’ ~ *Oxidation ’~ of products formed in hydroquinone states is proposed to occur via reaction with starting menogaril to produce menogaril hydroquinone. This latter step is the propagation step in the Fisher mechanismL3for nucleophilic trapping of quinone methides produced by reductive activation. The quinone methide 6 results from elimination of methanol from the menogaril hydroquinone state produced by reduction of menogaril by DHM-3 radical in one-electron steps8J4or from the chain propagation step described above. The byproduct of reduction by DHM-3 is 5,6-dihydro-3,5-dimethyl-5-(hydroxymethyl)1,4-oxazin-2-one @).I4 The chain mechanismL3and the catalytic effect of hydroquinones on the undesired tautomerization of 68explain the use of substoichiometric amounts of the reducing agent, DHM-3 dimer, to maximize the yield of 3. The competing process of dimerization is slow because of the low concentration of quinone methide during the course of the reaction. The successful coupling is also consistent with the higher reactivity with nucleophiles and the lower reactivity with electrophiles, such as the proton, at the 7-position anticipated for anthracyclines deprived of the O H group at the 11-position (menogaril,s aclacinomycin A,I2vL3and 11deoxydaunomycinL3J5).

(1 1) The material obtained from the reverse-phase column was partially protonated by the trifluoroacetic acid eluent. The N-Me2 signal appears as one large singlet and two small singlets; the two small singlets result from the fact that in the protonated form the methyls are diastereotopic and proton exchange in DMSO solvent is slow. Weak coupling with the N H is apparent in the COSY spectrum.

206, 1105.

+

R’ R‘

-

I

H, daunomycin OH, adriamycin

Similar reduction of menogaril in the presence of a large excess of 2’-deoxyadenosine did not result in adduct formation. The predominant species present after reaction were menogaril (1, (12) Kleyer, D. L.; Gaudiano, G.; Koch, T. H. J . Am. Chem. SOC.1984, (13) Ramakrishnan, K.; Fisher, J. J. Am. Chem. Soc. 1983, 105, 7187. Fisher, J.; Abdella, B. R. J.; McLane, K. Biochemistry 1985, 24, 3562. Ramakrishnan, K.; Fisher, J. J . Med. Chem. 1986, 29, 1215. (14) Gaudiano, G.; Koch, T. H. J. Org. Chem. 1987, 52, 3073. (15) Boldt, M.; Gaudiano, G.; Koch, T. H. J . Org. Chem. 1987,52,2146.

J . Am. Chem. SOC.1989, 1 1 1 , 8293-8294

34%), 7-deoxynogarol (4,43%), and bi(7-deoxynogarol-7-yl) (5, 9%). The unidentified minor product peaks with the exception of a peak at 7.2 min (5%) had the same retention times as the minor products from reduction in the presence of deoxyguanosine. The 7.2-min peak could not have been an adduct since an adduct would be considerably more polar than menogaril and would have eluted before menogaril. Its UV-vis spectrum also precluded it being an adduct. The result described here indicates for the first time that an anthracycline quinone methide transient can covalently bind to a nucleophilic site in a nucleic acid. Studies of menogaril metabolism show low levels of recovery of drug and its known metabolites.I6 Possibly, covalent binding to nucleophilic sites in proteins and/or nucleic acids through reductive activation is an explanation. The quinone methide from reductive activation of adriamycin or daunomycin bears an 11-hydroxy substituent and consequently is short lived and less likely to bind covalently to sites of lower n ~ c l e o p h i l i c i t y . ' ~ J ~InJ ~vivo modification of the 11-hydroxy group of these drugs prior to reductive activation might change the reactivity of their quinone methide states with nucleic acids. (16) Dcdion, P.; Egorin, M. J.; Engisch, K. L.; Bachur, N. R. Cancer Res. 1985, 45, 5352. Egorin, M. J.; Van Echo, D. A,; Whitacre, M. Y.; Forrest, A.; Sigman, L. M.; Engisch, K. L.; Aisner, J. Zbid. 1986, 46, 1513. (17) Kleyer, D. L.; Koch, T. H. J . Am. Chem. SOC.1984, 106, 2380.

Chemical Reaction Mechanisms at Unusually Low Temperatures: The Gas-Phase Reaction of C2H2+ H2

+

Michael Hawley and Mark A. Smith* Department of Chemistry, University of Arizona Tucson, Arizona 85721 Received March 27, 1989 We have recently developed a supersonic flow kinetic technique that allows reaction rate studies at extremely low collision ene r g i e ~ . ' - ~ We report the observation of gas-phase reactivity between the acetylene radical cation, C2H2+,and molecular hydrogen at collision energies of less than 0.5 meV (translational temperatures near 2 K). C2H2+

+ H2

-

C2H3+

+H

8293

3 t

T

2

1

1.8

I

1

I

2.0

2.2

2.4

Tl,,",

2.6

( K)

Figure 1. The observed translational temperature dependence of the experimentally measured rate coefficients for reaction 1. Temperatures are a neutral density weighted average of the perpendicular and parallel temperatures present in the jet over the range of an experiment. Mor,, = 267-269 kcal mo1-1.11,12All of these results have led to an acceptance of an endothermicity for reaction 1 of 2.0 kcal mol-I at 0 K. Our experiment is outlined below. Briefly, a 1-3% mixture of C2H2in H2 at 0.5-1.4 atm is formed into a free jet by expansion through a 0.03- or 0.05-cm orifice into a vacuum chamber. Downstream, a pulsed dye laser is used to selectively ionize C2H2 via resonant multiphoton ionization through the G In, u = 0 state of C2H2. This ionization technique allows selective production of C2H2+(or C2D2+in separate experiments using C2D2)in the ground vibrational state with 100% state purity.13 The photoions remain in translational equilibrium with the surrounding neutrals. Velocity and temperature slip under our expansion conditions are negligibleS5The properties of both the reactor and the expansion assure that all collisions occur at the extremely low energies characteristic of a free jet flow. Temporal monitoring of the chemical composition of the cold ion packet via mass spectrometry allows determination of an absolute reaction rate coefficient. The experimentally observed rate coefficient is actually an average over two distribution functions, the relative ion-neutral collision energy and the neutral rotational state distribution f ~ n c t i 0 n . l These ~ degrees of freedom are not in equilibrium in the jet, therefore,

(1)

Reaction 1 has been well studied at temperatures above 80 K and is found to increase in rate with increasing temperature.6-10 The observed rate coefficient at 300 K is 1.0 X lo-" cm3 s-l, and the reaction is not observed to occur at 80 K.6*7The reaction cross section is observed to increase with collision energy between 0.05 and 0.8 eV.8 Vibrational excitatiqn in the C2H2+also accelerates the rate at low e n e r g i e ~ . ~Appearance J~ potential measurements for C2H2+and C2H3+, from the photoionization of CzH4,have suggested that the heat of formation of C2H3' lies in the range (1) Mazely, T. L.; Smith, M. A. Chem. Phys. Lett. 1988, 144, 563-568. (2) Randeniya, L. K.; Zeng, X.K.; Smith, M. A. Chem. Phys. Lett. 1988, 147, 346-351. (3) Mazely, T. L.; Smith, M. A. J . Chem. Phys. 1988, 89, 2048-2062. (4) Randeniya, L. K.; Zeng, X. K.; Smith, R. S.; Smith, M. A. J. Phys. Chem., in press. ( 5 ) Hawley, M.; Mazely, T. L.; Randeniya, L. K.; Smith, R. S . ; Zeng, X. K.; Smith, M. A., submitted for publication in Znt. J . Mass Spectrom. Zon Processes. (6) Adams, N. G.; Smith, D. Chem. Phys. Lett. 1977, 47, 383-387. (7) Smith, D.; Adams, N. G.;Ferguson, E. E. In?.J. Mass Spectrom. Zon Processes 1984, 61, 15-1 9. (8) Turner, T.; Lee, Y. T. J. Chem. Phys. 1984, 81, 5638-5643. (9) Butrill, S. E.; Kim, J. K.; Huntress, W. T.; LeBreton, P.; Williamson, A. J . Chem. Phys. 1974, 61, 2122-2128. (IO) Honma, K.; Kato, T.; Tanaka, K.; Koyano, I. J. Chem. Phys. 1984, 81, 5666-5671.

0002-7863/89/1511-8293%01.50/0

Thef; are the appropriate distribution functions, k ( e j ) are the microcanonical rate coefficients, and e is the relative collision energy, whilej is the rotational quantum number for the diatomic reactant. It is well-known that H2 does not efficiently rotationally relax within a free jet, and the resulting rotational temperature is between 150-200 K under our expansion condition^.'^ Even at these temperatures, 98% of the H2 population resides in rotational levels w i t h j less than or equal to 2. At the same time, the translational distribution can be well represented by an average temperature in the range 1.8-2.5 K. These translational temperatures are averages over the parallel and perpendicular degrees of freedom in the jet. They are calculated by using the thermal conduction model of Klots16and the observed H, rotational tem(1 1) Chupka, W. A.; Berkowitz, J.; Refaey, K. M. A. J. Chem. Phys. 1969, 50, 1938-1941. (12) Stockbauer, R.; Inghram, M. C. J. Chem. Phys. 1975,62,4862-4870. (13) Orlando. T. M.; Anderson, S . L.; Appling, J. R.; White, M. G . J. Chem. Phys. 1987, 87, 852-859. (14) Only low-lying rotational levels of the ion are initially prepared. Since the cross section for exothermic rotational-to-translational energy transfer collisions should be very large, we assume that the rotational temperature of the ions is negligible and the distribution is collapsed to the rotationless ground state. (15) Gallagher, R. J.; Fenn, J. B. J. Chem. Phys. 1974, 60, 3492-3499. (16) Klots, C. E. J . Chem. Phys. 1980, 72, 192-197.

0 1989 American Chemical Society